Enhancing Health Fact-Checking with LLM-Generated Synthetic Data
By: Jingze Zhang , Jiahe Qian , Yiliang Zhou and more
Potential Business Impact:
Makes online health advice more trustworthy.
Fact-checking for health-related content is challenging due to the limited availability of annotated training data. In this study, we propose a synthetic data generation pipeline that leverages large language models (LLMs) to augment training data for health-related fact checking. In this pipeline, we summarize source documents, decompose the summaries into atomic facts, and use an LLM to construct sentence-fact entailment tables. From the entailment relations in the table, we further generate synthetic text-claim pairs with binary veracity labels. These synthetic data are then combined with the original data to fine-tune a BERT-based fact-checking model. Evaluation on two public datasets, PubHealth and SciFact, shows that our pipeline improved F1 scores by up to 0.019 and 0.049, respectively, compared to models trained only on the original data. These results highlight the effectiveness of LLM-driven synthetic data augmentation in enhancing the performance of health-related fact-checkers.
Similar Papers
A Case Study Exploring the Current Landscape of Synthetic Medical Record Generation with Commercial LLMs
Computation and Language
Makes fake health records work at any hospital.
Hybrid Fact-Checking that Integrates Knowledge Graphs, Large Language Models, and Search-Based Retrieval Agents Improves Interpretable Claim Verification
Computation and Language
Checks facts better by searching and thinking.
MedFact: Benchmarking the Fact-Checking Capabilities of Large Language Models on Chinese Medical Texts
Computation and Language
Helps doctors trust AI's medical advice.